Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Beyond Message Passing: A Semantic View of Agent Communication Protocols

Dun Yuan, Fuyuan Lyu, Ye Yuan, Weixu Zhang, Bowei He, Jiayi Geng, Linfeng Du, Zipeng Sun, Yankai Chen, Changjiang Han, Jikun Kang, Alex Chen, Haolun Wu, Xue Liu

Abstract

Agent communication protocols are becoming critical infrastructure for large language model (LLM) systems that must use tools, coordinate with other agents, and operate across heterogeneous environments. This work presents a human-inspired perspective on this emerging landscape by organizing agent communication into three layers: communication, syntactic, and semantic. Under this framework, we systematically analyze 18 representative protocols and compare how they support reliable transport, structured interaction, and meaning-level coordination. Our analysis shows a clear imbalance in current protocol design. Most protocols provide increasingly mature support for transport, streaming, schema definition, and lifecycle management, but offer limited protocol-level mechanisms for clarification, context alignment, and verification. As a result, semantic responsibilities are often pushed into prompts, wrappers, or application-specific orchestration logic, creating hidden interoperability and maintenance costs. To make this gap actionable, we further identify major forms of technical debt in today's protocol ecosystem and distill practical guidance for selecting protocols under different deployment settings. We conclude by outlining a research agenda for interoperable, secure, and semantically robust agent ecosystems that move beyond message passing toward shared understanding.

Beyond Message Passing: A Semantic View of Agent Communication Protocols

Abstract

Agent communication protocols are becoming critical infrastructure for large language model (LLM) systems that must use tools, coordinate with other agents, and operate across heterogeneous environments. This work presents a human-inspired perspective on this emerging landscape by organizing agent communication into three layers: communication, syntactic, and semantic. Under this framework, we systematically analyze 18 representative protocols and compare how they support reliable transport, structured interaction, and meaning-level coordination. Our analysis shows a clear imbalance in current protocol design. Most protocols provide increasingly mature support for transport, streaming, schema definition, and lifecycle management, but offer limited protocol-level mechanisms for clarification, context alignment, and verification. As a result, semantic responsibilities are often pushed into prompts, wrappers, or application-specific orchestration logic, creating hidden interoperability and maintenance costs. To make this gap actionable, we further identify major forms of technical debt in today's protocol ecosystem and distill practical guidance for selecting protocols under different deployment settings. We conclude by outlining a research agenda for interoperable, secure, and semantically robust agent ecosystems that move beyond message passing toward shared understanding.

Paper Structure

This paper contains 118 sections, 5 figures, 1 table.

Table of Contents

  1. Introduction
  2. Preliminary
  3. Definition of AI Agents
  4. From Classical MAS to Modern Agentic AI.
  5. Rise of Modern AI Agents.
  6. Motivation for Protocol Standardization.
  7. Agent Communication Protocols
  8. Historical Context.
  9. Modern Agent Protocols (2023–2025).
  10. Why Communication Protocols Matter.
  11. Human Communication and Interdisciplinary Motivation
  12. Foundations from Information Theory.
  13. Linguistic and Pragmatic Perspectives.
  14. Gap Between Human and Agent Communication.
  15. Motivating a Human-Inspired Taxonomy.
  16. ...and 103 more sections

Figures (5)

  • Figure 1: Human communication and agent communication can be compared through the same three-layer lens: communication, syntactic, and semantic alignment.
  • Figure 2: Overview of the proposed three-layer taxonomy for agent communication protocols, spanning communication, syntactic, and semantic concerns.
  • Figure 3: Illustrative semantic alignment failure and repair in A2A. After successful communication and syntactic parsing, the request remains semantically ambiguous because “Springfield” may refer to multiple destinations. The repair path detects the ambiguity, triggers a clarification exchange, updates shared context, and executes the task only after intent is aligned.
  • Figure 4: Internet of Agents architecture. Coral Servers coordinate heterogeneous agents through decentralized message routing and task orchestration. The system supports trusted interaction and cross-agent workflows, while higher-level alignment is mainly mediated through identity, contracts, and blockchain-based verification rather than explicit clarification mechanisms.
  • Figure 5: Decision workflow for selecting an agent communication protocol. Starting from deployment and trust-boundary requirements, the workflow routes common application settings to representative protocols, including MCP, ACP-AGNTCY, ACP-IBM, A2A, ANP, LMOS, Agora, Coral, WAP, and LOKA. The goal is to choose the simplest protocol that provides the required communication, syntactic, and semantic guarantees.